Model Based Identification and Fault Detection of an Active Suspension System

This paper presents an investigation on the performance of an active suspension system. After discussing a 4-DOF car suspension model, dynamic simulation of the system with an existing nonlinear classic controller called back-stepping method has been presented which had been previously introduced by Huang and Lin (2004). The model-based identification of the system is then preformed by the aid of the feed forward neural networks which are finally used for the corresponding fault detection process. Simulation results show that the proposed identification and fault detection approach is highly effective in evaluating the performance of an active suspension system.